Value of radiological diagnosis of skull fracture in the management of mild head injury:
meta-analysis
P A M Hofman, P Nelemans, G J Kemerink, J T Wilmink
Abstract
Objectives—Head injury is a common event. Most patients sustain a mild head injury (MHI), and management depends on the risk of an intracranial haemor- rhage (ICH). The value of a plain skull radiograph as a screening tool for ICH is controversial. The aim of this meta- analysis was to estimate and explain diVerences in reported sensitivity and specificity of the finding of a skull fracture for the diagnosis of ICH, in order to assess the value of the plain skull radiograph in the investigation of patients with MHI, and to estimate the prevalence of ICH in these patients.
Method—After a systematic literature search 20 studies were selected that re- ported data on the prevalence of ICH after MHI and/ or data on the diagnostic value of skull fracture for the diagnosis of ICH.
The mean prevalence of ICH weighted for the sample size was determined. The sen- sitivity and specificity of diVerent studies were combined using a summary receiver operator characteristic curve. Correlation analysis was used to determine factors that could explain the reported diVerences between studies.
Results—The weighted mean prevalence of ICH after MHI is 0.083. The potential for verification bias and the percentage of patients who had suVered loss of con- sciousness or post-traumatic amnesia were the most significant factors explain- ing interstudy diVerences in sensitivity and specificity. Based on studies wherein at least 50% of patients had a CT study of the brain, the estimated sensitivity of a radiographic finding of skull fracture for the diagnosis of ICH is 0.38 with a corresponding specificity of 0.95.
Conclusion—The plain skull radiograph is of little value in the initial assessment of MHI patients.
(J Neurol Neurosurg Psychiatry2000;68:416–422) Keywords: head injury; skull fracture; meta-analysis;
radiological diognosis
Head injury is one of the most common injuries and can be considered a silent epidemic. In the Western world it is one of the leading causes of disability, especially in the young population. The Head Injury Task Force of the National Institute of Neurologic Disor- ders has estimated that there are 2 000 000
cases of head injury in the United States annu- ally. In The Netherlands the estimated inci- dence of head injury is 0.14- 0.64%(Twijnstra 1998, personal communication), slightly less than the reported incidence in the United States. Most patients (80% to 90%) sustain a mild head injury (MHI) and do not need admission to hospital or complex health care. If these patients attend the emergency depart- ment of a hospital, almost all are sent home.
This, however, does not mean that MHI is a totally benign condition. An outcome study of patients who had a head injury suggested that patients with a low risk of dying—that is, patients with MHI—are at the greatest risk of inadequate diagnosis and treatment.1Consid- ering the many people aVected, little research has been done on the assessment and treat- ment of this category of patients. This is also reflected by the fact that management proto- cols for MHI are still under debate, which has led to considerable diVerences in strategies. In the past few years protocols have been published,2−4which might be seen as belonging to two diVerent schools: the North American and the European. In North America, the rou- tine use of CT for the radiological assessment of patients with MHI is currently under debate, whereas in Europe the use of a plain skull radiograph is disputed. The primary manage- ment goal in MHI is to identify those patients who are at risk of developing complications, specifically an intracranial haemorrhage (ICH) requiring surgery. Clinical assessment alone is inadequate for the detection of ICH,5 and radiological procedures are therefore used as additional screening tools. That patients with a skull fracture have an increased risk of intracranial haematoma is well known,6−8 but does this have practical significance? A skull fracture by itself has few clinical consequences, except in cases of a depressed skull fracture.
The potential clinical usefulness of radiological assessment for skull fracture depends on the ability to distinguish between patients with MHI with and without ICH.
To judge the usefulness of the diagnosis skull fracture, it is important to evaluate the sensitivity and specificity of this finding as a test for the presence or absence of ICH, and to determine the prevalence of ICH in patients with MHI. Unfortunately, sensitivity and specificity estimates reported in the literature show large variation. This may be because published studies diVer in design (prospective and retrospective approaches), patient selec- tion (admitted patients or patients seen at the Department of
Radiology, University Hospital Maastricht and the University Maastricht, PO Box 5800, 6200 AZ, Maastricht, The Netherlands P A M Hofman G J Kemerink J T Wilmink Department of Epidemiology P Nelemans Correspondence to:
Dr P A M Hofman, Department of Radiology, University Hospital Maastricht, PO Box 5800, 6200 AZ, Maastricht, The Netherlands
email [email protected] Received 26 April 1999 and in revised form
21 October 1999 Accepted 29 October 1999
emergency department), and inclusion criteria (based on Glasgow coma scale (GCS), loss of consciousness (LOC) or post-traumatic amne- sia (PTA)). Although ICH is mostly diagnosed by CT, in older studies it was diagnosed on the basis of clinical, operative, or postmortem findings. Comparison of the data is further complicated by possible diVerences in thresh- old for a positive test result (fracture or ICH) or by diVerences in technical instrumentation. In some studies considerable abnormalities may be required to be present before the test is declared positive, whereas others may require only a hint of abnormality. In the first case, sensitivity will be low and specificity high; in the second case sensitivity will be high and specificity low. The implication is that there is a trade oV of sensitivity against specificity between the studies, which needs to be taken into account in any method for combining results.
Given this diversity, it is presently not possi- ble to draw a conclusion about the value of radiography in detecting skull fracture in the management of patients with MHI, and for this reason we carried out a meta-analysis of published data, using correlation analysis to identify the most important sources of varia- tion in prevalence and diagnostic accuracy between studies, followed by use of the summary receiver operator characteristic curve (ROC) technique described by Moseset al,9to assess the eVect of these potential sources of variation, and to summarise reported sensitiv- ity and specificity estimates from the reviewed studies. Our aim was to assess the value of the diagnosis skull fracture for the diagnosis of ICH, and to summarise reported sensitivity and specificity estimates from reviewed studies.
We therefore tried to account for (part of) the diVerences in reported sensitivity and specifi- city of skull fracture for the diagnosis of ICH between studies. To be able to estimate the predictive value of the diagnosis skull fracture, the prevalence of ICH in patients with MHI was also estimated.
Material and methods
LITERATURE SEARCH STRATEGY AND DATA COLLECTION
A systematic search for relevant original publi- cations was conducted in Medline, Embase, and Current Contents from 1966 to 1998, using the following search keys: skull fracture, skull injury, skull radiography, skull trauma, skull films, (brain or head), (trauma or injury or injuries) and computed tomography (all subheadings). The articles were primarily selected on the basis of the title and the abstract. Additional references were obtained from the bibliographies of the original articles.
The full text of about 200 relevant articles was retrieved. Two sets of articles were selected, one set to estimate the diagnostic value of a finding of a skull fracture, and a second set to assess the prevalence of ICH in patients with MHI.
For the first set of papers the test under study is the plain skull radiograph for the determina- tion of the presence of an ICH. In those stud-
ies where no plain skull radiography was performed, CT data were used. Papers were included if they contained data on the diagnos- tic value of a finding of a skull fracture by plain radiograph or CT in patients who had MHI.
The second set of papers was selected for the assessment of the prevalence of ICH in MHI.
For these studies the standard of reference for diagnosis was the existence of ICH on CT.
Only a few studies fulfilled this strict criterion;
therefore we lowered the norm, and at least 50% of the patients needed to have undergone CT.
For the purpose of this study, MHI was defined as trauma to the head, with the patient having a Glasgow coma scale (GCS)10score of 13 to 15 on initial presentation. In the selected studies the diagnosis of ICH was made by CT.
If no CT was performed an uneventful recovery was considered a sign for the absence of an ICH. In some older studies angiography was used to diagnose ICH, and neurosurgical findings were used by some as well. An arbitrary minimum of 50 patients was required.
Studies with less patients will have a statisti- cally unreliable estimate of sensitivity, specifi- city, and prevalence. Studies with only paediat- ric or geriatric patients were not included. If the data permitted, multitrauma patients and referrals were excluded. A standard form was used to extract relevant data from the original articles on study and patient characteristics, and various test results (table 1).
ANALYSIS
Prevalence of ICH
Prevalence was defined as the percentage of patients in the study with a diagnosis of ICH.
Both mean prevalence weighted for sample size and unweighted mean prevalence were calcu- lated. The weighted mean was defined as11:
mean prev=∑(ùiprevi)_/∑ùi (1) withùi= 1/(previ(1-previ)/Ni) (2) Calculation of the true and false positive rate For evaluation of the diagnostic value of a skull fracture only the diagnosis of ICH was used, and not the report of a surgical intervention.
This choice was made firstly because the indi- cation for intervention diVered between insti- tutions and clear criteria were rarely given, and secondly because some investigators consid- ered the placement of intracranial pressure monitor devices as an intervention, whereas others excluded this procedure. In those
Table 1 Items extracted or derived from original studies
Study and patient characteristics Results
Publication year Number of skull radiographs Retrospective/prospective Number of CTs
Number of referrals True positive number (TP) Number of patients True negative number (TN) Age distribution False positive number (FP)
Mean age False negative number (FN)
Injury severity (GCS) Number of ICHs Percentage with LOC and/or PTA Number of interventions Focal neurology Number of deaths Other injuries
LOC=Loss of consciousness, PTA=post-traumatic amnesia.
studies where no plain skull radiography was performed, CT data on skull fractures were used.
From the collected data the number of true and false positive observations and the number of true and false negative observations were derived. True positive (TP) is defined as the finding of both a skull fracture and an ICH, false positive (FP) as a skull fracture without an ICH, true negative (TN) as the absence of both a skull fracture and an ICH, and false negative (FN) as the absence of a skull fracture in the presence of an ICH. Using these data the true positive rate (TPR=TP/(TP+FN)) and the false positive rate (FPR=FP/(FP+TN)) were calculated. The TPR equals the sensitivity and the FPR equals (1-specificity).
Correlation study to identify confounding diVerences between studies
The TPR and FPR are not independent.
Rather, there is a trade oVbetween the two, as is reflected in the ROC curve. Without an exact match of the study population and analysis characteristics, simply averaging these rates can be very misleading and does not provide a rep- resentative summary.12To determine the eVect of interstudy diVerences as mentioned in table 1, a correlation analysis was performed with parameter D, which is defined in the next sub- section, and which is a measure for how well the test discriminates between the population with and without ICH. The Spearman correla- tion test was used for this analysis.
Summary operator characteristic curve
For the analysis of TPR and FPR data, as found in the diVerent studies, we used a summary ROC (sROC) curve as described by Moseset al.9The analytical method is based on the principle that the sROC curve is conven- iently represented as a roughly straight line when logit TPR is plotted against logit FPR.
For statistical reasons, logit TPR-logit FPR (D) is modelled as a linear function of logit TPR+logit FPR (S).
S=logit (TPR)+logit (FRP) (3) D=logit (TPR)-logit (FRP) (4) with the logit defined as:
logit(x)=1n(x/(1-x) (5)
S is related to how often the test is positive and D is a direct measure of how well a test
discriminates between the population with an ICH and without an ICH, since:
D=1n(odds ratio) (6)
The odds ratio is a measure of association used in epidemiological studies. In diagnostic studies, the odds ratio is the odds of a positive test result in diseased patients divided by the odds of a positive test result in non-diseased patients. The higher the odds ratio, the better the test discriminates between patients with and without the disease of interest.13
To estimate the relation between S and D a linear model is fitted to the data:
D=áS+C (7)
C is a measure of the ability of the test to discriminate between diseased and non- diseased persons, and á is a measure of the extent to which D depends on the threshold for a positive test result. The higher the constant C, the better the discriminatory ability of the test. Using the fitted á and C, the relation between TPR and FPR can be transformed back into an sROC curve.
Equation 7 can be extended with further factors (F) in order to evaluate the influence of study and population characteristics on D:
D=áS+C+âF (8)
The goodness of fit was expressed by the square of the correlation coeYcient (R2) between the observed value of D and the predicted value of D. IfR2is 1, there is a perfect fit; ifR2is 0 there is no linear relation between the observed and the predicted value of D. The data analysis was performed using commer- cially available software (Microsoft Excel 5.0a and SPSS 6.1.1 for the PowerPC Macintosh).
Results
DESCRIPTION OF STUDIES
Twenty studies were identified that could be used to study the prevalence of ICH or the diagnostic value of the radiological detection of skull fracture for the diagnosis of ICH in adult patients with MHI. Thirteen of these studies contained data on the prevalence of ICH based on CT examinations.5 14−25Table 2 summarises the data from this group of studies. In 13 of the 20 studies, TPR and FPR of the finding of skull Table 2 Extracted data from the literature which were used to estimate the prevalence of ICH in MHI
First author (reference)
Publication
year Design ED/AD n
Severity (GCS)
N CTs
LOC/PTA (%)
Previous ICH (%)
Surgical interventions
(%) Deaths(%)
Livingston23 1991 R ED 111 14–15 111 82 13.5 0
Livingston25 1991 R ED 138 14–15 75 75 9.4 0.7
Mohanty14 1991 R AD 348 13–15 348 100 3.4 0 0
Rao24 1991 R 857 15 857 11.7 4.3
Harad5 1992 R ED 302 13–15 302 61 18.2
Shackford19 1992 R ED 2766 13–15 2166 100 16.9 4 0.2
Stein20 1992 R ED 1538 13–15 1538 100 12.9 3.8
Jeret18 1993 P ED 712 15 712 100 9.4 0.3 0.1
Borczuk22 1995 R ED 1448 13–15 1448 80 6.3 0.8 0
Dunham21 1996 R ED 2032 13–15 2032 100 6.3 0.4 0.1
Ingebrigtsen17 1996 R AD 91 13–15 88 100 8.8 0 0
Holmes16 1997 P ED 264 14 264 100 12.1 1.5 0
Miller15 1997 P ED 2143 15 2143 100 5.1 0.2 0
ED=Emergency department; AD=admitted patients R=retrospective; P=prospective.
fracture in predicting ICH could in principle be calculated.17 19−22 25−32Although two studies included patients with moderate and severe head injury, these studies were nevertheless included because most patients had MHI (over 90%).29 30It seems unlikely that the small pro- portion of patients with moderate and severe head injury in these studies could have a major impact on the conclusions of the meta-analysis.
In five studies no plain skull radiography was done and in these studies CT data on skull fractures were used to assess the relation with
ICH. Nine studies were retrospective; the oth- ers were prospective. Table 3 summarises the characteristics of these studies. Note that there is overlap between the two groups of tables 2 and 3: six studies were used for both analyses.
PREVALENCE OF ICH AND CORRELATIONS
The mean prevalence of ICH after MHI was 0.1 (95% confidence interval (95% CI) 0.02–0.18, range 0.03–0.18) and the weighted mean prevalence was 0.083 (95% CI 0.03–
0.13, table 2).
DIAGNOSTIC ACCURACY
The sensitivity (TPR) of the finding of skull fracture in predicting ICH ranged from 0.13 to 0.75 and the specificity (1-FPR) from 0.91 to 0.995. The mean D of all studies was 3.35, and the mean sensitivity was 0.50, corresponding to a specificity of 0.97 on the sROC (figure).
Studies with a high TPR tended to have a higher FPR, but the fit of the sROC curve to the observed pairs of sensitivity and specificity values was poor (R2=0.08). Therefore, the dif- ferences in discriminatory ability between studies cannot be explained by diVerences in diagnostic thresholds for positive test results.
Consequently, an alternative explanation was needed for the variation in sensitivity and spe- cificity. Spearman rank correlation analysis showed that the percentage of patients with LOC/PTA and the percentage of patients who had undergone CT was significantly correlated with D (table 4). A model based on equation 8, which included (besides C and S) a factor rep- resenting the percentage of patients with LOC/
PTA fitted the data better (R2=0.73). Addition of a factor representing the percentage of patients who underwent CT resulted in an even better fit (R2=0.81). This confirms that diVer- ences in patient selection and the percentage of patients in whom the diagnosis was verified by CT were important sources of variation between studies.
Sensitivity and specificity are not invariant to the population under study, and often they will depend on patient characteristics—for exam- ple, patient selection. In clinical studies this is often a reflection of clinical practice. For example, a study with patients admitted for MHI is likely to have a more severely injured population than a study with only emergency department patients. We considered patient selection as an important source of variation Table 3 Extracted data on the value of the radiological diagnosis of skull fracture in the assessment of ICH in MHI. These figures were used to estimate the sensitivity and specificity of the diagnosis skull fracture for the diagnosis of ICH. The sROCs of these data are shown in figure 1
First author (reference)
Publication
year ED/AD n Design Severity Modality
No of skull radiographs NCTs
LOC/P TA
(%) CT (%) TPR FPR
Royal College28 1981 ED 5850 P 13–15 X 5850 <50 <50 0.7500 0.0133
Dacey27 1986 AD 610 P 13–15 X 583 68 100 11 0.7222 0.0817
Kraus26 1988 AD 2402 R 13–15 X 2402 <50 <50 0.3876 0.0854
Gorman29 1987 ED 12395 P 0–15 X 5484 15 <50 0.7273 0.0362
Masters30 1987 7035 P 0–15 X 4068 <50 <50 0.5556 0.0174
Livingston5 1991 ED 138 R 14–15 X 71 75 75 54 0.4286 0.0313
Shackford19 1992 ED 2766 R 13–15 X 423 2166 100 78 0.6082 0.2708
Stein20 1992 ED 1538 R 13–15 CT 1538 100 100 0.6034 0.0649
Borczuk22 1995 ED 1448 R 13–15 CT 1448 80 100 0.1319 0.0052
Dunham21 1996 ED 2032 R 13–15 CT 2032 100 100 0.2734 0.0436
Gomez31 1996 ED 2484 R 13–15 CT 1784 187 28 7.5 0.5581 0.0106
Ingebrigtsen17 1996 AD 91 R 13–15 CT 88 100 97 0.2500 0.0750
Arienta32 1997 ED 9830 R 13–15 X 6724 969 10 9.8 0.5484 0.0083
ED=Emergency department AD=admitted patients; R=retrospective; P=prospective; X=plain skull radiography
sROC curve of the diagnostic value of the radiological finding of a skull fracture for the diagnosis of ICH in MHI.
1.0
0.8
0.7
0.6 0.9
0.5
0.4
0.2 0.3
(26)
(17) (21)
(19)
(25) (30, 31)
(20) (27)
Group 1, studies with <50% patients LOC/PTA and <50% CT
Group 2, studies with >50% patients LOC/PTA and >50% CT
Study with 100% patients LOC/PTA and 11% CT
(28)(29)
0.1
0.0 0.5
FPR
TPR
0.0 0.1 0.2 0.3 0.4
Table 4 Results of Spearman correlation analysis between potential confounding factors and D
Predictor CoeYcient p Value
Percentage of patients with LOC/PTA (selection bias) −0.6792 0.011 Percentage of CT scans (verification bias) −0.6472 0.017 Skull radiograph compared with CT for fracture diagnosis 0.2535 0.403 Prospective studiesvretrospective studies 0.4900 0.089
Adult populationvall ages 0.0976 0.751
See equation 4 for the definition of D.
between the estimates of sensitivity and specifi- city, and the percentage of patients with LOC/
PTA was the most significant selection crite- rion.
Selection of patients who underwent CT depending on the plain skull radiography results, or on patient characteristics, will result in verification bias, also called work up bias. We used the percentage of patients who underwent CT as a measure of verification bias. The lower the percentage of patients in whom the diagno- sis of ICH was verified by CT, the higher the potential for verification bias. Although not explicitly mentioned in the studies, it is very likely that the decision to perform a CT was based on the patient assessment and/or the skull radiography findings.
The percentage of patients with LOC/PTA was strongly correlated with the potential for verification bias. Two groups of studies were formed. The first group contained the studies in which fewer than 50% of patients had LOC/
PTA and fewer than 50% of patients had CT (group 1). This group had thus a high potential for verification bias. The second group con- tained studies for which both percentages were higher than 50% (group 2). There is only one study that did not fit in either of the two groups.27The sROC curve fitted to the data for group 1 using equation 7, showed that in this population a relatively high TPR was reached at a low FPR (figure). The sROC curve of data for group 2 was lower than the sROC curve of group 1. The mean D in group 1 and group 2 was 4.3 and 2.4 (p=0.016), respectively. Sum- mary values for sensitivity and specificity are not directly available from the analysis, but estimates can be read oVthe sROC curve. The mean sensitivity was 0.59 and 0.38, with corre- sponding specificities of 0.98 and 0.95, for groups 1 and 2, respectively.
Discussion
In this meta-analysis we investigated the value of radiological assessment for skull fracture in the diagnosis of ICH in patients with MHI and analysed the prevalence of ICH in this category of patients.
Despite the high incidence of MHI, relatively few well designed prospective studies on the management of MHI have been published. All studies found by our literature search were biased to a lesser or greater extent. Firstly, the percentage of patients with a history of LOC/PTA varied considerably, resulting in patient selection bias, and secondly, the percentage of patients in whom the diagnosis of ICH was verified by CT was highly variable, resulting in a potential for verification bias in most studies. In earlier studies only a small percentage of patients underwent cranial CT, and even nowadays patients with a GCS score of 15 but no a history of LOC/PTA seldom undergo CT. In older studies cerebral angio- graphy, and operative and postmortem findings were used to establish the diagnosis of ICH.
The mean prevalence of ICH in patients with MHI was 0.10, the range 0.03–0.18, with a weighted mean of 0.083. The percentage of patients with LOC/PTA was relatively high in
the studies that were used to derive this preva- lence. In the studies with a low percentage of patients with LOC/PTA, fewer patients under- went CT (higher potential for verification bias). A high prevalence of ICH has also been found in studies including only patients with a GCS score of 15 and LOC/PTA.18 24
A strong potential for verification bias—that is, few CT scans—leads to an overestimation of the sensitivity. Patients with a negative skull radiograph will not undergo CT, so patients with false negative results will have a higher chance of remaining undetected.33 This bias could oVer an explanation for a mean sensitiv- ity of 0.59 for group 1 (less than 50% CT), compared with a sensitivity of 0.38 for group 2 (over 50% CT). The unverified negative test results (no skull fracture) were assumed to have no ICH, and this will result in an overestima- tion of the specificity.33The data corroborate this: the specificity of 0.98 for group 1 (higher potential for verification bias) is higher than the specificity of 0.95 for group 2 (lower potential for verification bias). Patient selection bias and verification bias were strongly associated in the studies investigated. This made it possible to distinguish one group of studies with both a low percentage of patients with LOC/PTA and few undergoing CT, and a second group with a high percentage of patients with LOC/PTA and relatively many undergoing CT. Because verifi- cation bias aVects the sensitivity,33the sensitiv- ity of the radiological finding “skull fracture”
for the diagnosis ICH is most reliably obtained from studies with a low verification bias (group 2). In that group the mean sensitivity was 0.38 with a corresponding specificity of 0.95.
It should be kept in mind that an sROC curve diVers from a traditional ROC curve.
The ROC curve describes the relation between sensitivity and specificity in a single popula- tion, with a changing threshold. The sROC curve results from fitting a smooth line to data points representing pairs of sensitivity and spe- cificity values from diVerent studies and thus diVerent populations. Therefore, the area under the curve, as a measure of overall diagnostic accuracy, cannot be determined for the sROC curve, whereas it can for the traditional ROC curve.9
By combining the results for sensitivity, spe- cificity, and prevalence, it is possible to calculate the positive predictive (PPV=TP/
(TP+FP)) and negative predictive value (NPV=TN/(TN+FN)) of the radiological de- tection of skull fracture for the diagnosisof ICH. With a sensitivity of 0.38 and a specificity of 0.95, as found for group 2, and a prevalence of 0.083, the PPV is 0.41 and the NPV is 0.94.
This means that if there is a skull fracture, the probability of an ICH is about 4.9 times higher Table 5 Findings to be expected for a fictitious population of 1000 patients with MHI, characterised by an ICH prevalence of 0.083, in combination with a test sensitivity of 0.38 and a specificity of 0.95
ICH+ ICH−
Skull fracture+ 32 46 77 PPV=41%
Skull fracrure− 51 871 923 NPV=94%
83 917 1000
than before testing. A plain skull radiograph increased the probability of no ICH from 92%
to 94%. What these figures mean in clinical practice is illustrated in table 5 for a fictitious group of 1000 patients. The most important conclusion of this review is that a positive skull radiograph does not predict an ICH with certainty, although the risk is definitely in- creased. More importantly, at a sensitivity of 0.38, a plain skull radiograph does not provide much extra information and cannot be used for ruling out the diagnosis of ICH.
The findings of this review contradicts data from the literature. Of the 735 patients who had an ICH in the 13 studies, only 322 (44%) had a skull fracture. Therefore, the claim that 80% of patients with ICH have a skull fracture8 is not valid. Moreover, at a prevalence of 0.083, the probability of ICH in patients with MHI and a skull fracture is about five times higher than in patients without a skull fracture. This is in contradiction to the 41-fold increased risk mentioned by Mendelow et al.34 There was a strong potential for verification bias in that study, because ICH was verified in only a few patients. This may explain the high sensitivity of 0.75 and the high relative risk of ICH in patients with a skull fracture in that study. Fur- thermore, in the studies included in this meta- analysis the prevalence of ICH in patients with MHI presenting at an emergency department was in the order of 0.03 to 0.10, rather than the reported value of 0.003.34
A few points need to be discussed. The first point concerns the use of both plain skull radiography and CT (in five studies) to detect skull fracture. The plain skull radiograph is considered to be more sensitive for the detection of calvarial skull fracture than CT, whereas CT is more sensitive for the detection of skull base fractures. In the light of other dif- ferences between the studies, we considered that this possibly not fully equivalent sensitivity was acceptable. A second point concerns the use of the diagnosis of ICH as the gold stand- ard, instead of intervention or clinical course.
The existence of ICH is of clinical importance as an indicator of the severity of the trauma and as a guideline for rehabilitation.35 36It may very well be that the many ICHs that went undetec- ted until recently is in part responsible for the high incidence of post-concussional syndrome in patients with MHI.37
In all studies, the radiologist’s report of the skull radiograph was used, whereas in daily practice the emergency physician or resident assesses the radiographs, and management is based on these initial findings. Thillainayagam et alshowed that up to 10% of skull fractures are missed by less experienced physicians,38 who usually see most patients with MHI in many institutions. This will decrease the sensi- tivity of the skull radiograph even further.
The estimated mean prevalence of ICH after MHI was 0.083. The two most significant fac- tors explaining the interstudy diVerence in reported sensitivity and specificity of the exist- ence of a skull fracture for the diagnosis ICH are the percentage of patients with LOC/PTA and the potential for verification bias. We con-
clude that the plain skull radiograph has no place in the assessment of MHI in adult patients. The question is not whether the detection of a skull fracture ever assists in the detection of ICH, but whether this is eVective.
Our analysis shows that the plain skull radiograph was ineVective as a screening tool for patients with MHI: only slightly more than one third of ICH were detected in this way. The low sensitivity implies that if a skull fracture is not seen on plain skull radiography, the diagnosis of ICH still cannot be ruled out. If patient selection increases the likelihood of ICH, CT becomes the modality of first choice.
Data from the literature also suggest that some patients with MHI and a GCS score of 15 do not require any imaging. Two studies described a subpopulation of MHI patients with a GCS score of 15 and no LOC/PTA or any other neurological symptoms.30 32None of these patients had an ICH, and no interven- tions were needed. Patients with a GCS score of 15 and LOC/PTA, and patients with a GCS score of 13 and 14, require either observation, CT, or both.
1 Klauber MR, Marshall LF, Luerssen TG,et al. Determi- nants of head injury mortality: importance of the low risk patient [[comments]].Neurosurgery1989;24:31–6.
2 Stein CS. Management of minor closed head injury.Neuro- surgery Quartely1996;6:108–15.
3 Hutchinson PJA, Kirkpatrick PJ, Addison J,et al. The man- agement of minor traumatic brain injury.J Accid Emerg Med1998;15:84–8.
4 Servadei F, Ciucci G, Loroni L, et al. Diagnosis and management of minor head injury: a regional multicenter approach in Italy.J Trauma1995;39:696–701.
5 Harad FT, Kerstein MD. Inadequacy of bedside clinical indicators in identifying significant intracranial injury in trauma patients.J Trauma1992;32:359–61.
6 Servadei F, Ciucci G, Morichetti A,et al. Skull fracture as a factor of increased risk in minor head injuries. Indication for a broader use of cerebral computed tomography scanning.Surg Neurol1988;30:364–9.
7 Galbraith S, Mendelow AD, Jennett B. Skull X-rays [[letter]].Lancet1981;ii:1350.
8 Evans KT. The radiologist’s dilemma [[editorial]]. Br J Radiol1977;50:299–301.
9 Moses LE, Shapiro D, Littenberg B. Combining independ- ent studies of a diagnostic test into a summary ROC curve:
data-analytic approaches and some additional considera- tions.Stat Med1993;12:1293–316.
10 Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale.Lancet1974;ii:81–4.
11 Bradburn MJ, Deeks JJ, Altman DG. Metan: an alternative meta-analysis command.Stata Technical Bulletin1998;44:
4–15.
12 Irwig L, Tosteson ANA, Gatsonis C,et al. Guidelines for meta-analyses evaluating diagnostic tests.Ann Intern Med 1994;120:667–76.
13 Sackett DL, Haynes RB, Guyatt GH,et al.Clinical epidemi- ology. A basic science for clinical medicine.London: Little, Brown, 1991.
14 Mohanty SK, Thompson W, Rakower S. Are CT scans for head injury patients always necessary?J Trauma1991;31:
801–4.
15 Miller EC, Holmes JF, Derlet RW. Utilizing clinical factors to reduce head CT scan ordering for minor head trauma patients.J Emerg Med1997;15:453–7.
16 Holmes JF, Baier ME, Derlet RW. Failure of the Miller cri- teria to predict significant intracranial injury in patients with a Glasgow coma scale score of 14 after minor head trauma.Acad Emerg Med1997;4:788–92.
17 Ingebrigtsen T, Romner B. Routine early CT-scan is cost saving after minor head injury.Acta Neurol Scand1996;93:
207–10.
18 Jeret JS, Mandell M, Anziska B,et al. Clinical predictors of abnormality disclosed by computed tomography after mild head trauma [[comments]].Neurosurgery1993;32:9–15.
19 Shackford SR, Wald SL, Ross SE,et al. The clinical utility of computed tomographic scanning and neurologic examina- tion in the management of patients with minor head injuries [[comments]].J Trauma1992;33:385–94.
20 Stein SC, Ross SE. Mild head injury: a plea for routine early CT scanning [[comments]].J Trauma1992;33:11–3.
21 Dunham CM, Coates S, Cooper C. Compelling evidence for discretionary brain computed tomographic imaging in those patients with mild cognitive impairment after blunt trauma.J Trauma1996;41:679–86.
22 Borczuk P. Predictors of intracranial injury in patients with mild head trauma.Ann Emerg Med1995;25:731–6.
23 Livingston DH, Loder PA, Koziol J,et al. The use of CT scanning to triage patients requiring admission following minimal head injury.J Trauma1991;31:483–7.
24 Rao R. Discussion on Mohantyet al. Are CT scans for head injury patients always necessary?J Trauma1991;31:801–4.
J Trauma1991;31:804.
25 Livingston DH, Loder PA, Hunt CD. Minimal head injury:
is admission necessary?Am Surg1991;57:14–7.
26 Kraus JF, Nourjah P. The epidemiology of mild, uncompli- cated brain injury.J Trauma1988;28:1637–43.
27 Dacey RG Jr, Alves WM, Rimel RW,et al. Neurosurgical complications after apparently minor head injury. Assess- ment of risk in a series of 610 patients. J Neurosurg 1986;65:203–10.
28 Costs and benefits of skull radiography for head injury. A national study by the Royal College of Radiologists.Lancet 1981;ii:791–5.
29 Gorman DF. The utility of post-traumatic skull xrays.
Archives of Emergency Medicine1987;4:141–50.
30 Masters SJ, McClean PM, Arcarese JS,et al. Skullxray examinations after head trauma. Recommendations by a multidisciplinary panel and validation study.N Engl J Med 1987;316:84–91.
31 Gomez PA, Lobato RD, Ortega JM,et al. Mild head injury:
diVerences in prognosis among patients with a Glasgow coma scale score of 13 to 15 and analysis of factors associ- ated with abnormal CT findings.Br J Neurosurg1996;10:
453–60.
32 Arienta C, Caroli M, Balbi S. Management of head-injured patients in the emergency department: a practical protocol.
Surg Neurol1997;48:213–9.
33 Begg CB, McNeil BJ. Assessment of radiologic tests: control of bias and other design considerations.Radiology1988;
167:565–9.
34 Mendelow AD, Teasdale G, Jennett B,et al. Risks of intrac- ranial haematoma in head injured adults.British Medical Journal Clinical Research Edition1983;287:1173–6.
35 Williams DH, Levin HS, Eisenberg HM. Mild head injury classification.Neurosurgery1990;27:422–8.
36 Hsiang JN, Yeung T, Yu AL,et al. High-risk mild head injury.J Neurosurg1997;87:234–8.
37 Brown SJ, Fann JR, Grant I. Postconcussional disorder:
time to acknowledge a common source of neurobehavioral morbidity.J Neuropsychiatry Clin Neurosci1994;6:15–22.
38 Thillainayagam K, MacMillan R, Mendelow AD, Brookes MT, Mowat W, Jennett B. How accurately are fractures of the skull diagnosed in an accident and emergency department.Injury1987;18:319–321.
